Flashcards
Network Layer
Flashcards do conjunto (68)
- 1What is the network layer concerned with?- Transmission of data over the internetwork (interconnected networks)
- Getting packets from source to destination (end-to-end transmission) - 2What is used to join a network?A gateway
- 3How can networks be different?- Protocol
- Addressing Systems
- Multicasting
- Datagram/Virtual Circuit
- QoS
- Error Handling
- Flow Control
- Congestion Control
- Security
- Parameters - 4What are collection-less networks also known as and how do they work?Datagram Networks
Data split into packets (datagrams) and are sent and routed independently - 5What are connection oriented networks also known as and how do they work?Virtual-Circuit Networks
Connection must be established between sender and receiver
All information follows same routes - 6Advantages of Datagram Networks+ Routes don't need to store path data
+ No need to establish connection between host and destination
+ Balance of traffic throughput - 7Disadvantages of Datagram Networks- Packet must hold source and dest. address in header
- QoS is complex
- Difficult to control congested path(s)
- Each packet routed independently - 8Advantages of Virtual Circuit Networks+ Packet contains a short VC number
+ Routing a sequence of packets performed on same route
+ Guarantee of QoS
+ Avoids congestion - 9Disadvantages of Virtual Circuit Networks- Routers must research memory for routing tables
- Virtual circuit connection is required
- Vulnerability issues - 10What is the main networking protocol used in the internet?The Internet Protocol (IP)
- 11What are the characteristics of the IP?- Unreliable protocol
- Connectionless delivery
- Responsible for transporting packets/datagrams
- No re-ordering facility - 12What are the two types of communication offered by the IP?1. Unicasting - Between 1 sender and 1 receiver
2. Multicasting - 1 to many communication - 13What does TCP stand for and what are its characteristics?- Provide reliable communication over unreliable internetwork
- Acknowledgement of data arrival
- Flow control using a numbering sys.
- Error control
- Congestion control - 14TCP is most widely used for...File transfer
- 15Who created the TCP/IP model?Gerf & Kahn
- 16What are the sections of the TCP/IP model?Application
Transport
Internet
Host-to-Network - 17What layers of the ISO OSI are contained within the TCP/IP- Application, Presentation and Session all contained within Application in TCP/IP
- Transport the same
- Network = Internet
- Data Link and Physical contained within Host-to-Network - 18How many bits are used in an IPv4 IP address?32 bits
- 19How many bits are used in an IPv6 IP address?128 bits
- 20How many classes can an IPv4 IP address be divided into and what are they?5
A, B, C, D, E - 21IPv4 Class A - Who was it designed for? Number of networks? Number of Hosts?Large organisations w/ many hosts and routers
Networks: 2^7
Hosts: 2^24 - 22IPv4 Class B - Who was it designed for? Number of networks? Number of Hosts?Mid-sized organisations w/ thousands of attached hosts/routers
Networks: 2^14
Hosts: 2^16 - 23IPv4 Class C - Who was it designed for? Number of networks? Number of Hosts?Smaller Organisations
Networks: 2^21
Hosts: 2^8 - 24IPv4 Class D - Who was it designed for? Number of networks? Number of Hosts?Reserved for multicasting (datagram sent to multiple hosts at same time)
Networks: 2^0
Hosts: 2^28 - 25IPv4 Class E - Number of networks? Number of Hosts?Networks: 2^0
Hosts: 2^28 - 26What is a mask?A 32 bit number comprised of 1's and 0's that help us to split IP address into netid and hostid
- 27What does CIDR stand for?Classless Interdomain Routing
- 28What is the main problem associated with classful addressing?The Three Bears Problem
A - Too large for almost any organisation
B - Too large for most that were assigned B
C - Too small
D - Overestimated
E - Hardly used - 29How frequently is classful addressing used?Rarely!
The three bears problem is responsible - 30Why was classless addressing introduced and what does it use rather than classes?To give more organisation access to the internet
Works with address blocks rather than classes - 31What are the restrictions of classless addressing?- Addresses must be in contiguous blocks
- No. of addresses must be a power of 2
- First address must be evenly divisible by number of addresses - 32What is a mask?A 32 bit number, n leftmost bits are 1's and rightmost bits are 0's
- 33x.y.z.t /n What does this represent?x.y.z.t gives an IP address
/n gives the mask - 34Why was IPv6 created and what were the goals associated with this version?Created as IPv4 addresses will run out
- Reduce routing table size
- Support more hosts
- Allow faster packet processing
- Improved Security - 35What are the 3 main strategies associated with transitioning from IPv4 to IPv6?- Dual Stack
- Tunneling
- Header Translation - 36IPv4 - IPv6: TunnelingUsed when 2 machines want to communicate with each other where the packets must pass though and IPv4 network.
IPv6 packet is encapsulated within an IPv4 packet - 37IPv4 - IPv6: Dual StackNetwork layer split into 2 separate blocks running IPv4 and IPv6 simultaneously
- 38IPv4 - IPv6: Header TranslationUsed when most of internet is IPv6, header translated from IPv6 to IPv4 if receiver does not understand IPv6 yet
- 39What are subnets?Subnetworks
- 40Subnet CalculationSee Sheet/Do an example
- 41What is the main protocol used in the network layer of the TCP/IP model?IP
- 42What are the protocols associated with Address Mapping?ARP - Address Resolution Protocol
RARP - Reverse Address Resolution Protocol
BOOTP - Bootstrap Protocol
DHCP - Dynamic Host Configuration Protocol - 43What are the protocols associated with Error Reporting & Net Management/DiagnosticsICMP - Internet Control Message Protocol
- 44What are the protocols associated with Multicast Membership Management?IGMP - Internet Group Management Protocol
- 45How is an IP address obtained from a MAC address?- Static Mapping
- Dynamic Mapping - 46How does static mapping work?- A table is stored on each machine
- Mappings from logical to physical address - 47Limitations of Static Mapping- Machine could change its NIC(Network Interface Controller), producing a new physical address
- Mobile devices roam between networks, resulting in physical address changes - 48How does dynamic mapping work?A machine knows either/both logical and physical addresses, it can use a protocol to find the other one
- 49What does ARP stand for and how does it work?Address Resolution Protocol
- ARP query packet sent from host and broadcasted on the network
- Intended recipient recognises IP address and responds to ARP query w/ ARP response packet
- Once sender receives this packets can be sent - 50What is the structure of an ARP packetSee Lecture 10 slide 76
- 51What does RARP stand for and what does it do?Reverse Address Resolution Protocol
- Resolves a logical address to physical address
e.g. a diskless station is booted, An organisation doesn't have enough IP addresses and must assign on demand - 52What is the main issue with RARP?A RARP server is required on each network
- 53What does BOOTP stand for and what is its purpose?Bootstrap Protocol
- A replacement for RARP
- Application layer protocol
- Uses a relay agen to allow packets to be forwarded from LAN - 54What does DHCP stand for and what is its purpose?Dynamic Host Configuration Protocol
DHCP has static and dynamic address locations
Allows manual and auto configuration (unlike DHCP) - 55DHCP: What are the differences between static and dynamic?Static - Emulates BOOTP
Dynamic - DHCP has a db containing a pool of logical IP addresses
When DHCP client requests IP address, server finds used IP from pool and assigns for a limited time - 56DHCP: When is dynamic used frequently?Moving to and from different networks
- 57DHCP: How is an IP address obtained?- New host broadcasts DHCP DISCOVER packet
- DHCP intercepts this packet and send a unicast packet to DHCP server - 58DHCP: What is it called when an IP is assigned for a fixed time?Leasing
- 59DHCP: How can a fixed IP time be extended?Before lease expires, host requests an extension
- 60What does ICMP stand for and when is it used?Internet Control Message Protocol
- Used to provide unreliable connectionless datagram delivery - 61What 2 types of messages does ICMP produce?- Error-reporting messages
- Query messages - 62How big is the header for a ICMP packet?8 bytes
- 63ICMP: Error-reporting messages- Destination Unreachable: A host cannot delivery a datagram
- Source Quench: Lack of flow control in IP, leads to router congestion
- Time Exceeded: Datagrams travelling in a loop
- Parameter Problem: Ambiguities in header of datagram
- Redirection: Mechanism to update host routing tables - 64ICMP: Query Messages- Echo request and reply: Used for diagnostics by admin
- Timestamp request and reply: Used for RTT measurements
- Address-mask request and reply: Obtain corresponding address mask for the network
- Router solicitation and advertisement: Host requests routing info from routers on network - 65ICMP: PING- Source send ICMP echo-request
- ping sends packet with ICMP sequence No.
- Each packet gives a RTT
- TTL given value, if hops exceeds this packet timeout
- sends 56 bytes of data (84 w/ header)
- Deliver stats after interruption - 66ICMP: TracerouteFill if time
- 67What does IGMP stand for and what is its purpose?Internet Group Management Protocol
- Can use unicast (1:1) and multicast(1:M) - 68How many implementations of IGMP are there?3
IGMPv1
IGMPv2
IGMPv3